In the communications field it is often advantageous to transmit signals representative of the human voice without transmitting the entire voice signal. Several vocoder algorithm are available for analyzing the human voice so that a representative signal greatly reduced in bandwidth and information can be transmitted. When the signal is received the human voice is synthesized, or reproduced, from the received signal.
In general, the vocoder algorithms require a tremendous amount of circuitry or the bit rate of communications between circuits is extremely high so that a relatively large amount of power is required.